The Third-Person Effect in Communication W. Phillips Davison The ...

The Third-Person Effect in Communication
W. Phillips Davison
The Public Opinion Quarterly, Vol. 47, No. 1 (Spring, 1983), 1-15.
Stable URL:
http://links.jstor.org/sici?sici=0033-36228 198321%2947%3A1%3C1%3ATTEIC%3E2.O.C0%3B2-9
The Public Opinion Quarterly is currently published by Oxford University Press.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you
have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and
you may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.j stor.org/journals/oup.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or
printed page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of
scholarly journals. For more information regarding JSTOR, please contact support@jstor.org.
http://www.jstor.org/
Sun Mar 6 03:27:28 2005

The Third-Person Effect
in Communication
W. PHILLIPS DAVISON
IN 1949 or 1950, while combing through cartons of U.S. Marine Corps
documents from World War 11, a young historian at Princeton Uni-
versity came across a series that piqued his curiosity. He stepped
across the corridor and described his find to an even younger
sociologist:
"You're supposed to know something about public opinion. What
do you make of this? There was a service unit consisting of Negro
troops with white officers on Iwo Jima Island in the Pacific. The
Japanese learned about the location of this unit and sent planes over
with propaganda leaflets. These leaflets stressed the theme that this
was a white man's war and that the Japanese had no quarrel with
colored peoples. They said, more or less, 'Don't risk your life for the
white man. Give yourself up at the first opportunity, or just desert.
Don't take chances.' The next day that unit was withdrawn."
"Why do you find this so interesting?'asked the sociologist.
"Because I can't find any evidence that the propaganda had an
Abstract A person exposed to a persuasive communication in the mass media sees this
as having a greater effect on others than on himself or herself. Each individual reasons:
"I will not be influenced, but they (the thud persons) may well be persuaded." In some
cases, a communication leads to action not because of its impact on those to whom it is
ostensibly directed, but because others (third persons) think that it will have an impact
on its audience. Four small experiments that tend to support this hypothesis are
presented, and its complementary relationship to a number of concepts in the social
sciences is noted. The thud-person effect may help to explain various aspects of social
behavior, including the fear of heretical propaganda by religious leaders and the fear of
dissent by political rulers. It appears to be related to the phenomenon of censorship in
general: the censor never admits to being influenced; it is others with "more im-
pressionable minds" who will be affected.
W. Phillips Davison is Professor of Journalism and Sociology at Columbia Univer-
sity. The author wishes to thank Robert L. Cohen and two anonymous referees for
reading an earlier version of this article and making helpful comments.
Public Opinion Quarterly Val. 47:l-15 @ 1983 by the Trustees of Columbia University
Published by Elsevier Science Publishing Co., Inc. 0033-362X/83/0047-1/$2.50

2 W. PHILLIPS DAVISON
effect on the troops at all. But it sure had an effect on the white
officers. The leaflets seem to have caused a substantial reshuffle of
personnel." 1
The sociologist mumbled something about probable guilt feelings on
the part of the white officers and a tendency of the military to prefer
solutions that involve physical action. But he couldn't escape the
feeling that something else was involved.
A few years later, in the course of investigating the role of the West
German press in the formation of Bonn's foreign policy, the
sociologist had occasion to ask a series of journalists how much
influence they thought newspaper editorials had on the thinking of
their readers (Davison, 1957). One of the replies given frequently was
along the following lines: "The editorials have little effect on people
like you and me, but the ordinary reader is likely to be influenced
quite a lot." Since evidence to support such a judgment could not be
located, this line of inquiry was eventually abandoned, but the researcher
remained impressed with the extent to which many journalists
were convinced that editorials had an effect on other people's
attitudes, while discounting the effect on people like themselves.
Some time after that, the sociologist became involved in the local
phase of a national election, serving as a volunteer for his preferred
candidate's organization. Two days before the election a leaflet supporting
the rival candidate appeared in his mailbox. He was impressed
with its quality. It would undoubtedly swing a lot of votes. Some
counteraction would have to be taken. Without thinking further, he
procured a pile of political literature from his own party's local office
and spent the rest of the day distributing it door to door.
Informal postelection analyses (no systematic studies were conducted
at the local level) suggested that neither set of propaganda
materials had exerted much influence on the voters. It was as though
a page had been taken out of The People's Choi~e.~ The sociologist
The historian in question, Jeter Isely, died tragically shortly thereafter. The book
resulting, in part, from his research includes no direct reference to this incident; indeed,
it makes no reference at all to Japanese leaflets, but it does contain the following
passage: "About 200 of the enemy slipped through into the rear zones. There they were
liquidated by the Fifth Pioneer Battalion and other service troops, many of them
Negroes. The corps shore party commander was 'highly gratified with the performance
of these colored troops. , . . While in direct action against the enemy for the first time
. . . they conducted themselves with marked coolness and courage' " (Isely and Crowl,
1951:SOO). It would seem probable that the leaflets were dropped following this en-
gagement.
It is scarcely necessary to remind readers that this study of the 1940 election
campaign found that very few voters were converted by political propaganda
(Lazarsfeld et al., 1944).

TPIIRD.PE;RSON EFFECT IN COMMUNICATION 3
(who is identical with the writer of this article) began to ask himself
why he had assumed that the rival candidate's leaflet would be so
effective.
These personal experiences, and probably others that have been
forgotten, led to the formation of a proposition that, for want of a
better label, may be called the "third-person effect hypothesis." In its
broadest formulation, this hypothesis predicts that people will tend to
overestimate the influence that mass communications have on the
attitudes and behavior of others. More specifically, individuals who
are members of an audience that is exposed to a persuasive communi-
cation (whether or not this communication is intended to be persua-
sive) will expect the communication to have a greater effect on others
than on themselves. And whether or not these individuals are among
the ostensible audience for the message, the impact that they expect
this communication to have on others may lead them to take some
action. Any effect that the communication achieves may thus be due
not to the reaction of the ostensible audience but rather to the behav-
ior of those who anticipate, or think they perceive, some reaction on
the part of others.
The phenomenon under consideration has been called the "third-
person effect" because third persons are involved from two different
observational standpoints. In the view of those trying to evaluate the
effects of a communication, its greatest impact will not be on "me" or
"you," but on "themw-the third persons. From the standpoint of a
propagandist or other persuasive communicator, on the other hand,
the third persons are those who are in some way concerned with the
attitudes and behavior of the ostensible audience. Indeed, the prop-
agandist may try to manipulate the behavior of these third persons by
apparently seeking to influence someone else,
This second definition of the "third person" may have been in the
minds of the Japanese strategists who arranged to have leaflets
dropped over the black service units on Iwo Jima. They may not have
expected the leaflets to have an effect on the troops themselves, but
were instead trying to goad the white military command into taking
the action that it apparently did take in fact-namely, to withdraw the
service units.
Imputation of such reasoning to Japanese military propagandists is
supported by the fact that British and American psychological warfare
in Europe made use of a very similar tactic. The History of the
Psychological Warfare Division, Supreme Headquarters, Allied Ex-
peditionary Force (Bad Homburg, Germany, 1945), tells us about
Operation Huguenot-a project for undermining the efficiency of the
German Air Force by suggesting that German flying personnel were

4 W. PHlLLIPS DAVISON
deserting in their machines to the Allied side. Planting such sug-
gestions was not difficult. It was known that Allied radio broadcasts
were systematically monitored by the German government and that
monitoring reports were distributed to all high political and military
officials. Hints about desertions from the Luftwaffe could include, for
example, a "slip" by an announcer indicating that a plane officially
reported as being shot down had in fact landed safely in England. It
could be assumed that at least some of these hints would be picked up
by alert radio monitors in Berlin. The Psychological Warfare Division
history tells us:
The dividends from this operation were expected not so much in the actual
number of desertions as in the effect of the countermeasures which the
German authorities would be induced to take against flying personnel . . .
sharpening up of anti-desertion measures and instructions to field police to
keep a suspicious eye on everyone-a course which would have serious
effects on morale. Also, the promotion of officers on account of reliability
rather than efficiency (p. 53).3
It seems probable that practical persuaders throughout the cen-
turies have been aware of this use of the third-person effect. Lovers,
certainly, have frequently tried to influence the behavior of the loved
one by seeming to direct their attentions to someone else.
Four Small Experiments
During the past several years, the writer has made a series of minor
efforts to test one variant of the third-person effect hypothesis: that
an individual who is exposed to a persuasive communication via the
mass media will see this communication as having a greater effect on
other people than on himself or herself. All these tests have been
conducted with small groups under informal conditions. Elegant ex-
periments they were not. Nevertheless, care was taken to insure that
the groups did not contaminate each other and that their members did
not supsect that a particular hypothesis was being tested. Even
though no single experiment can be regarded as particularly impres-
sive in itself, the results all tend to confirm the hypothesis. Taken
together, they are reasonably convincing, at least to this writer.
The first trial was conducted with the help of a good-natured group
of 33 graduate students taking a course on mass communication at
According to an academic student of Allied psychological warfare in World War 11,
a number of desertions were claimed as a result of Operation Huguenot, even though
the principal purpose of the undertaking had not been to encourage desertions (Lerner,
1949:268).

THIRD-PERSON EFFECT IN COMMUNICATION 5
Columbia University in 1978. It was just after the New York State
election of that year and also just after a strike that had shut down the
three major New York City newspapers. A questionnaire included
items about both the election and the strike, and was divided into two
sections, one including "questions about New Yorkers in general,"
and the other "a few questions about your own experiences." One of
the items in the first section read as follows:
As you probably know, Governor Carey repeatedly called on Mr. Duryea [the
Republican challenger] to make his income tax returns public, and used Mr.
Duryea's failure to do so as a major campaign theme. About how much
influence do you think this had on the way New Yorkers voted in the
gubernatorial election? Please indicate this by making a mark at the appropri-
ate point on the scale below.
The scale ran from 0 (No Influence at All) to 7 (Very Great Influ-
ence).
At the end of the second section of the questionnaire, there was the
following item:
And how about Governor Carey's emphasis on Mr. Duryea's failure to make
his income tax returns public? If you had been a New York voter (or if you
actually were a New York voter), how much influence do you think this
would have had (or actually did have) on your vote in the gubernatorial
election? Please indicate this by making a mark at the appropriate point on
the scale below. [The same scale was used as in the first question.]
When the scores for New Yorkers in general were tabulated, it was
found that they had an arithmetic mean of 3.4, falling close to the
middle of the influence scale. The scores for personal influence had a
mean of only 2.26, indicating that, as predicted, the respondents
evaluated the persuasive communication as having a greater effect on
others than on themselves. Since the standard deviations of both
distributions were huge, however, the replies given by each respon-
dent were scored so as to ascertain how many individuals did in fact
see the communication as having more influence on New Yorkers in
general than on themselves. The results showed that about half the
respondents perceived the effect on others to be greater than on the
self, and that very few evaluated the effect as being greater on self
than on others:
More influence on New Yorkers in general 48%
More influence on self 6
Same influence on public and on self 36
No answerlno opinion 9
Another experiment designed to confirm or disconfirm the third-
person effect hypothesis was embodied in a small poll on mass media

6 W. PHILLIPS DAVISON
and socialization that was administered to 25 graduate students in the
spring of 1981. This time, the effect of communications on self was
asked about before an evaluation of the effect on others was re-
quested. Respondents were asked how often their parents had read to
them when they were small children, about how old they had been
when they learned to distinguish television commercials from program
content, and a few other personal questions. Then they were asked:
And how about attitudes toward commercial products? Did exposure to TV
influence you to ask your parents to buy things you otherwise wouldn't have
wanted?
The questionnaire went on to request observations about "other
people, especially other people's children." One item was:
Does exposure to TV cause kids to ask their parents to buy them things they
otherwise wouldn't want?
Again, the hypothesis seemed to be confirmed, although the small
number of respondents and the distribution of replies, as shown in
Table 1, leave adequate room for those who prefer to remain skepti-
cal.
In discussing the results, several respondents pointed out, quite
correctly, that the two questions had different wordings and that their
comparability was further diminished by the fact that they dealt with
different time periods. One noted that perhaps advertising techniques
had improved during the past 20 years to such an extent that today's
children are indeed more influenced by television commercials than
those of the 1960s.
The hypothesis was tested also during the primary campaigns prior
to the 1980 presidential election. A group of 25 adults, about equally
divided between those under 30 and over 30, who were attending a
lecture series at the Museum of Broadcasting in New York, were
asked a number of,questions regarding the upcoming election and the
role of media in it. Again, the questions about the effects of com-
Table 1. Television's Ability to Make Children Ask Parents to Buy Commercial Products
Self Influenced by Children of Others
TV as Child Influenced by TV
Quite a lot 32% 68%
Some 24 28
Not very much 20 4
Not at all 16 0
Didn't watch as a child 8 0

THIRD-PERSON EFFECT IN COMMUNICATION 7
munications on self and other were widely separated in the question-
naire, and subsequent discussion showed that no respondent had
become aware that a particular hypothesis was being tested. The
question on self read as follows:
Let's assume that you are planning to vote in the upcoming presidential
election. Would you say that your voting intention has been influenced by the
results of the New Hampshire primary?
Two pages later, there were two additional questions about the effects
of the New Hampshire primary on voters in general, although the
phraseology made it possible, perhaps mistakenly, to interpret the
questions as applying to other factors in additon to voting:
How much effect do you think the results of the New Hampshire primary will
have on the political fortunes of Ronald Reagan?
And how about Jimmy Carter? How much effect will the results of the New
Hampshire primary have on his political fortunes?
The results obtained for these three questions are shown in Table 2.
Again, some comfort is given to supporters of the third-person effect
hypothesis.
In this case, of course, the persuasive communications were not
necessarily of a propagandistic nature, and more likely consisted of
news reports. And the questions were not strictly comparable.
A fourth trial, again at the Museum of Broadcasting, followed the
same general format but asked a different group of respondents to
evaluate the effect on their own votes and the votes of people in
general of charges that Ronald Reagan would pursue a "hawkish
foreign policy. About twice as many respondents felt that other
people would be influenced more than they would be themselves.
Two further experiments, conducted in the fall of 1981 and the
spring of 1982, showed results very similar to the ones already de-
scribed.
Table 2. Effects of New Hampshire Primary on Personal Voting Intentions and on Political
Fortunes of Two Major Candidates
Will Influence Will Influence Will Influence
Own Intention Reapan's Fortunes Carter's Fortunes
-
Quite a lot 0% 52% 32%
A little 24 24 48
Not at all 72 20 20
Not sure 4 4 0

W. PHILLIPS DAVISON
Third-Person Effect in the Literature
Numerous scholars seem to have noted what has here been called
the third-person effect, but none so far as I know has paused to
comment on it. For example, the hypothesis seems to be supported
by several studies of the "Roots" television series. First aired in
January 1977, this eight-part dramatization of Alex Haley's story
about his forebears' painful progress from Africa, through slavery,
into the post-civil war period in the United States attracted over 130
million Americans to one or more episodes-the largest television
audience for any program up to that time. Even before all episodes in
the series had been shown, research organizations had started efforts
to gauge its effects on the public. Several of these research projects
were able to compare expected effects with observed effects. Most
respondents predicted that the series would have substantial impact
on the attitudes of both blacks and whites. The white reaction was
expected to be increased tolerance and sympathy; blacks were ex-
pected to be angry and to show bitterness and hostility (Howard et
al., 1978). When asked for their own reactions, however, substantial
pluralities of both blacks and whites reported that a feeling of sadness
was the principal effect of watching "Roots." Two researchers com-
mented that, in general, the program did not have the widespread
effects on racial attitudes attributed to it by observers (Hur and
Robinson, 1978). An analyst who reviewed five of the "Roots" studies
concluded that they provided little evidence of change in actual racial
attitudes, even though these had been widely expected (Surlin, 1978).
Somewhat similar observations were made in the course of an
investigation of American attitudes toward Jews during the period
following World War 11. It was hypothesized that the brutal persecu-
tion of European Jews by the Nazis and the strong support of the
Allied cause by Jews everywhere might have caused Americans gen-
erally to look with greater favor on their Jewish fellow citizens. Two
separate surveys by the Opinion Research Corporation in 1945
showed that this was not the case. Nearly four out of five respondents
in both surveys said that the mass killings of Jews in Europe had
caused no change in their attitudes toward Jews in the United States.
When asked in one of the surveys, however, over half of the respon-
dents said that they expected other people's attitudes to change, in
either a favorable or an unfavorable direction (Stember, et al.,
1966:142-43). Many influences in addition to the mass media were
involved here, of course, but newspaper and radio were certainly the
principal channels through which most people learned about the per-
secution of European Jewry.
A number of scholars have speculated that "experts" are particu-
larly likely to overemphasize the effects of the media. A journalist

THIRD-PERSON EFFECT IN COMMUNICATION 9
turned political scientist attributes this tendency especially to students
of politics and to communication theorists, explaining that this may be
because they are isolated from the actual operation of media organi-
zations (Diamond, 1978).
Another reason for misevaluation of media effects on the part of
experts is suggested by a study of the role of the press in community
conflict. In connection with an interview with an expert on a locally
controversial issue the authors note: "This expert's view is typical, in
the sense that it includes the belief that media affect people in general
but not the individual who has specialized expertise" (Tichenor et al.,
1980: 130).
A similar observation was made in the course of a study of percep-
tion of public opinion by decision makers in the field of nuclear
power. Many of these experts expressed the belief that the public was
being misled by biased coverage of nuclear power in the mass media,
since most people did not have access to good sources of technical
information (Cohen, 1982).
In a sense, we are all experts on those subjects that matter to us, in
that we have information not available to other people. This informa-
tion may not be of a factual or technical nature; it may have to do
with our own experiences, likes, and dislikes. Other people, we rea-
son, do not know what we know. Therefore, they are more likely to
be influenced by the media.
The literature contains fewer instances of cases in which people's
behavior has been influenced by their expectations that the media
would persuade others. Nevertheless, examples can be found, espe-
cially in connection with voter behavior. The third-person effect
seems to have been at work in the 1978 gubernatorial primary election
in Maryland, where "a reform candidate breezed by the incumbent
and his principal challenger while neither of them was looking" (Hol-
lander, 1979:405). According to repeated polls by two different orga-
nizations, only about 5 percent of the electorate expressed the inten-
tion of voting for this reform candidate, although he had an excellent
reputation. Then he received the editorial endorsement of the Balti-
more Sun and Evening Sun. A poll in progress when the endorsement
was made showed the reform candidate as being the choice of 4
percent before the "Sunpapers" editorial stand and 11 percent after-
wards. His performance in subsequent polls continued to improve and
he won a narrow victory in the election. As a public opinion re-
searcher in Baltimore put it, "The newspaper endorsement made
Hughes a plausible candidate and the voters did the rest" (Hollander,
1979:407).
To interpret these data as supporting the third-person effect hy-
pothesis one has to assume that the newspaper endorsements did not

10 W. PHILLIPS DAVISON
change the attitudes of many people toward Hughes; they rather
changed expectations about the support he would receive from
others. Individuals may have reasoned: "I am going to vote for him
because the newspapers have probably convinced other people of
something I already know-namely, that he is the best candidate-
and therefore he has a good chance of winning."
Such an interpretation was given by the campaign manager of
former mayor Carl Stokes of Cleveland, who mentioned at a Colum-
bia University seminar on public communication that the endorse-
ment by the Cleveland Plain Dealer was one of the major reasons for
Stokes' victory in the 1967 mayoralty race.
"Do you think the Plain Dealer's support really changed many
people's opinions," he was asked.
"I don't think it changed any," was the reply. "But it convinced
some individuals and organizations that he had a chance; so they
started sending in campaign contributions."
A similar observation about campaign contributions was made by
the press secretary of Senator Fred Harris, who ran for the Demo-
cratic nomination in the primaries prior to the 1976 presidential elec-
tion:
But in order to raise that kind of money dispersed among twenty states, then
you need national media exposure. You need it because people do judge by
national media exposure as to whether the campaign is serious or not and,
believe me, they hesitate before they give money. . . . They're going to Wait
until they see Fred's smiling face on national television (quoted in Arterton,
1978:9).
It is probable that advertisers and marketers are aware of the
action-inducing potential of the third-person effect, although I have
not noted references to this in the research literature. The frequently
used appeal, "Buy yours while the supply lasts," certainly suggests
that others will be persuaded by the advertising and that one had
therefore better make a purchase promptly.
Relationships to Other Phenomena
Media bias is frequently perceived in situations where it is clearly
absent or where it is present to a very limited extent. For example, in
the 1972 presidential campaign, many of those who preferred
McGovern thought their newspaper was giving more attention to
Nixon; and many who were for Nixon thought McGovern was receiving
more exposure in the same medium. The same judgments
were made with regard to television news, although less frequently.
Yet analysis of the media in question showed that they were giving

THIRD-PERSON EFFECT IN COMMUNICATION 11
fairly evenhanded treatment to the two major candidates (Mendelsohn
and O'Keefe, 1976: 148, 150).
A somewhat similar observation was made in the study of percep-
tions of public opinion by decision makers in the nuclear energy
sector that was referred to above. The decision makers frequently
complained that their field of operation received negative and sensa-
tional coverage by the mass media, and that this was a major reason
why many people opposed nuclear power. At the same time, oppo-
nents of nuclear power showed themselves to be equally dissatisfied
with news coverage in this area, zind accused the media of an estab-
lishment bias (Cohen, 1982).
One possible explanation for the fact that people on both sides of an
issue can see the media as biased against their own point of view is
that each observer assumes a disproportionate effect will be achieved
by arguments or facts supporting the "wrong" side of the issue.
Others (the third persons), the observer reasons, will be unduly im-
pressed by these facts or arguments; they do not have the information
that enables me to form a correct opinion. It is probable that, from the
point of view of partisans, balanced media presentation would require
a sharp tilt toward the "correct" side of the issue. This would com-
pensate for the intellectual frailty of third persons and would, ac-
cording to a partisan, ensure that the media achieved a truly balanced
presentation.
But, if the third-person effect hypothesis is correct, why are not the
facts and arguments on the "correct" side as well as the "wrong" side
seen as having a disproportionate effect on others? Perhaps the mate-
rial on the "correct" side is not seen as persuasive at all; it is merely a
statement of the obvious and therefore cannot be expected to have an
impact on attitudes.
Pluralistic ignorance, and the misperception of others' attitudes in
general, may also involve the third-person effect, at least in some
cases. If individuals assume that they are virtually alone in holding
particular attitudes and expectations, not knowing that many others
privately share them (Merton, 1968:431), it may be because they
assume others have been brainwashed by the mass media. Indeed, the
tendency to perceive the media as being biased toward the "wrong"
side of an issue, combined with the tendency to impute persuasive-
ness to the media insofar as others are concerned, creates a strong
presumption that the attitudes of other people on any controversial
issue that is in the focus of public attention will be widely misper-
ceived.
Empirical studies of pluralistic ignorance of which I am aware do
not offer much support for the above proposition. But neither do they
disconfirm it with any vigor. These studies have not treated public

12 W. PHILLIPS DAVISON
communication as a major variable to be built into their research
designs (for example, Schanck, 1932 and 1938; Colombotos et al.,
1975; Fields and Schuman, 1976; O'Gorman and Garry, 1976).
Schanck does, however, remark in connection with one controversy
in Elm Hollow that "continuous advocacy of a position by a minority
leads them (the majority) to a belief that this stand may be fairly
universal in their group" (Schanck, 1938:93).
To test the notion that the third-person effect plays a part in the
creation of pluralistic ignorance one could, for example, determine
whether misperception of others' attitudes is more likely to occur on
issues that have been extensively discussed in the mass media than on
issues that are discussed mainly in primary groups. The data pre-
sented by Fields and Schuman suggest that this might well be the
case, in that pluralistic ignorance appears to be greater in regard to
issues such as race relations (discussed widely in the media) than in
regard to issues involving ethics or morals (discussed more in primary
groups).
It is also possible that public communication diminishes mispercep-
tion of others' attitudes among people in part of the media audience
while it increases misperception among some. In this connection, the
concept of reference groups may prove useful in explaining the third-
person effect. Are people "like me" or "different from me" seen as
being more affected by persuasive messages? Or is the degree of
similarity not a relevant factor? If perceived congruity of others'
attitudes and values with one's own is a factor in the selection of
normative reference groups (Singer, 198 1 : 73), then one would expect
there to be little exaggeration in the perceived impact of a communi-
cation on members of such groups. On the other hand, the importance
of not overlooking a possible change in attitude on the part of a
significant other might make one assume, conservatively, that some
reference group members had indeed been affected. There is plenty of
room for research and speculation.
The third-person effect is probably involved in the "spiral of si-
lence" which, according to a recent theory about the formation of
public opinion, leads those on one side of an issue to express their
opinions with more and more volume and confidence, while those on
the other side of the issue tend to fall silent (Noelle-Neumann, 1980).
In particular, exaggerated perceptions of the effects of mass media
election propaganda on others would help to explain the situations in
which polls show that respondents think Party A will win an election
even though a majority of the respondents retain their intention of
voting for Party B. Each person may reason: I haven't been influ-
enced by this widely publicized nonsense, but they probably have
been.

THLRD-PERSON EFFECT IN COMMUNICATION 13
Thus, in both the 1965 and 1972 election campaigns in West Ger-
many, expectations as to which of the major parties would win the
election changed by as much as 15 percentage points during the
months prior to election day, while at the same time voting intentions
remained fairly stable (Noelle-Neumann, 1980: 15-17). At the last
minute, again consistent with the third-person effect hypothesis, un-
decided voters opted disproportionately in favor of the party which
seemed to be attracting the greater number of supporters. The rea-
soning of at least some of these late deciders was probably along the
following lines: I don't find much difference between the parties, but
the fact that others seem to be persuaded by the arguments or image
of Party A probably means that this is the better party. A bandwagon
effect was created.
As in the case of observed media bias, the question arises why the
impact of Party B's propaganda is not exaggerated as much as that of
Party A's propaganda. Perhaps there is simply more of the latter. This
was the case with some of the West German elections that were
studied. But it seems likely that other factors, presently unknown,
interact with the third-person effect to produce the observed result.
The Third-Person Effect in Our Lives
Fluctuations in the stock market are not infrequently accounted for
by reference to rumors or news reports-perhaps that a subcommittee
of the House of Representatives is considering a tax on widgets or that
the Ambassadors from Israel and an unnamed Arab state have been
seen sharing a taxi from Kennedy Airport to the United Nations
headquarters. The reasoning seems to be that these reports will cause
others to sell (or buy) certain categories of shares; therefore I will sell
(or buy) in order to anticipate their action.
In times when supplies of consumer goods are irregular, there are
always some people who will rush to the stores the moment they hear
reports of any possible shortage. If you ask them why, the answer is
likely to be that they are concerned about the effects of these reports
on other people. They want to stock up before the hoarders remove
all goods from the shelves.
When news stories about the possibly dangerous effects of aerosol
on the earth's atmosphere began to appear, according to an item in
the New York Times (September 16, 1975), manufacturers of products
sold in aerosol cans changed quickly to spray and squeeze containers.
Of course, there could have been a number of explanations for such
behavior (expectations of government regulations, the lower cost of
squeeze containers, etc.), but one possibility is that manufacturers

14 W. PHILLIPS DAVISON
expected stories about the dangers of fluorocarbon propellants to turn
the public against aerosol cans and were thus providing another
example of the third-person effect at work.
The phenomenon of censorship offers what is perhaps the most
interesting field for speculation about the role of the third-person
effect. Insofar as faith and morals are concerned, at least, it is difficult
to find a censor who will admit to having been adversely affected by
the information whose dissemination is to be prohibited. Even the
censor's friends are usually safe from pollution. It is the general
public that must be protected. Or else, it is youthful members of the
general public, or those with impressionable minds. When Maryland's
State Board of Censors, which had been filtering smut from motion
pictures since 1916, was finally allowed to die in June 1981, some of
its members issued dire forecasts about the future morals of Maryland
and the nation (New York Times, June 29, 1981). Yet the censors
themselves had apparently emerged unscathed. One of them stated
that over the course of 21 years she had "looked at more naked
bodies than 50,000 doctors," but the effect of this experience was
apparently more on her diet than on her morals. "I had to stop eating
a lot of food because of what they do with it in these movies," she is
quoted as having told the Maryland Legislature.
Throughout history, heretical doctrines and political dissidence
have aroused concern, sometimes terror, among priests and poten-
tates. How much of this apprehension and the resulting repression
was due to the third-person effect? It certainly must have played a
role, and probably has accounted for a grisly percentage of the
world's suffering and horror. Exaggerated expectations about the
effects of dissident communications have caused countless people to
be incarcerated, tortured, and killed. Even today, prisons in au-
thoritarian and totalitarian countries contain people whose crime is
alleged to be "propaganda against the state," or "spreading destruc-
tive rumors ."
Why are exaggerated expectations about the effects of communica-
tions on others so common? Do they occur in response to all
categories of persuasive communications, or only certain categories?
Or, is it possible that we do not overestimate effects on others so
much as we underestimate effects on ourselves?
References
Arterton, F. Christopher
1978 "Campaign organizations confront the media-political environment." In James

THIRD-PERSON EFE"ECT IN COMMUNICATION 15
David Barber (ed.), Race for the Presidency. Englewood Cliffs, N.J.:
Prentice-Hall.
Cohen, Robert L.
1982 "The perception and evaluation of public opinion by decision makers: civilian
nuclear power in the United States." Ph.D, dissertation, Columbia University.
Colombotos, John, Corrinne Kirchner, and Michael Millman
1975 "Physicians view national health insurance." Medical Care 13:369-96.
Davison, w.. Phillips
1957 "The mass media in West German uolitical life." In Hans Speier and W. P.
Davison, eds., West German ~eadeishi~ and Foreign ~olict Evanston, Ill.:
Row Peterson.
Diamond, Edwin
1978 Good News, Bad News. Cambridge, Mass.: MIT Press.
Fields, James M., and Howard Schuman
1976 "Public beliefs about beliefs of the public." Public Opinion Quarterly
40:427-49.
Hollander, Sidney
1979 "On the strength of a newspaper endorsement." Public Opinion Quarterly
43:405-7.
Howard, John, George Rothbart, and Lee Sloan
1978 "The response to 'Roots': a national survey." Journal of Broadcasting
22:279-88.
Hur, Kenneth K., and John P. Robinson
1978 "The social impact of 'Roots'." Journalism Quarterly 55:19-24.
1sely;Jeter A., and A. Crow1
1951 The U.S. Marines and Amphibious War: Its Theory and Its Practice in the
Pacific. Princeton, N.J.: Princeton University Press.
Lazarsfeld, Paul F., Bernard Berelson, and Hazel Gaudet
1944 The Peode's Choice. New York: Duell, Sloan & Pearce.
Lerner, Daniel
1949 Sykewar: Psychological Warfare Against Germany, D-Day to VE-Day. New
York: George W. Stewart.
Mendelsohn, ~arold; and Garrett J. O'Keefe
1976 The People Choose a President. New York: Praeger.
Merton, Robert K.
1968 Social Theory and Social Structure. New York: Free Press.
Noelle-Neumann, Elisabeth
1980 Die Schweigespirale. Munchen and Zurich: Piper Verlag.
O'Gorman, Hubert, and Stephen L. Garry
1976 "Pluralistic ignorance-a replication and extension." Public Opinion Quarterly
40:449-58.
Schanck, Richard L.
1932 "The community and its groups and institutions." Psychological Monographs
43, No. 2.
1938 "Test-tube for public opinion: a rural community." Public Opinion Quarterly
29-95.
Singer, Eleanor
1981 "Reference groups and social evaluations." In Morris Rosenberg and Ralph H.
Turner (eds.), Social Psychology: Sociological Perspectives. New York: Basic
Books.
Stember, Charles Herbert, et. al.
1966 Jews in the Mind of America. New York: Basic Books.
Surlin, Stuart U.
1978 " 'Roots' research: a summary of findings." Journal of Broadcasting
22:309-20.
Tichenor, Phillip J., George A. Donohue, and Clarice N. Olien.
1980 Community Conflict and the Press. Beverly Hills, Calif.: Sage.